Staff Software Engineer - Video Performance - (bay Area Only)

Canva Canva · Enterprise · San Francisco, CA +1 · Information Technology

Staff Software Engineer focused on video performance optimization, including ML-powered effect pipelines, C++ engine, shaders, and graphics APIs. The role involves designing and building observability systems, profiling, and collaborating with AI engineers to improve system architecture for parallelism and asynchronous processing.

What you'd actually do

  1. Lead end-to-end performance strategy across the video creation lifecycle — from timeline scrubbing and real-time preview through to export and rendering. You’ll identify bottlenecks and drive measurable improvements to the editing experience.
  2. Profile and optimize critical paths in our C++ native engine, GLSL/HLSL shaders, and ML-powered effect pipelines to reduce latency and memory footprint.
  3. Design and build observability systems including internal profiling tools and automated regression suites that capture key performance metrics (FPS, jank, memory usage, thermal behavior) across Web, iOS, and Android.
  4. Define and own performance metrics that directly reflect user-perceived fluidity and creative flow — ensuring performance remains a first-class product feature.
  5. Collaborate across teams with Video Engine and AI engineers to evolve system architecture for improved parallelism, cache locality, and asynchronous processing.

Skills

Required

  • C++
  • profiling and diagnostics
  • multimedia fundamentals
  • video codecs
  • container formats
  • frame-accurate seeking and composition
  • telemetry analysis
  • systems thinking
  • graphics APIs (OpenGL, Metal, Vulkan)
  • CPU/GPU architectures
  • memory hierarchies
  • SIMD

Nice to have

  • Rust
  • hardware acceleration (Metal, Vulkan, WebGPU)

What the JD emphasized

  • ML-powered effect pipelines
  • performance
  • performance
  • performance
  • performance

Other signals

  • optimize critical paths in our C++ native engine, GLSL/HLSL shaders, and ML-powered effect pipelines
  • Design and build observability systems including internal profiling tools and automated regression suites
  • Leverage hardware acceleration using platform-specific APIs (Metal, Vulkan, WebGPU)